• by Joe Mandese  @mp_joemandese, July 12, 2011

Before you read this story, I'd like you to take a mental leap and imagine that, despite this contrived start, it turns out to be pretty insightful, and important enough to make you want to save it, or maybe even pass it on to a friend or colleague. In the old days, you might have torn it out of this magazine, or made a duplicate on a copier and forwarded it through the postal service. But, more likely than not, now you would simply copy and paste the link to the online version into an email, tweet, personal blog or social media page. That would be apt, because this story fundamentally is about that simple human behavior - the act of copying and pasting links, text, pictures or other content from a Web page - and saving them or passing them along to others. And even more importantly: what publishers, marketers and agencies can learn from tracking, aggregating and analyzing that behavior.

 It's a behavior that may be taken for granted but is probably as prevalent as many of the online industry's most common behavioral metrics - clicks, searches, liking, friending, tweeting, etc. Until recently, however, it had been virtually ignored by Madison Avenue, though it says a lot about users' intentions regarding the content they read, listen to and watch. Not to mention how much they care about it, or the brands associated with it.

We might never have realized this if an accidental discovery hadn't led the management team of Web analytics firm Tynt to conclude that some powerful data could be unearthed by following the trail of copy-and-paste behavior. The data is powerful enough to make Tynt one of the 10 largest sources of data about online users - and, most likely, the largest one you've never heard of.

The epiphany occurred when Tynt chairman Allan MacKenzie's son was diagnosed with meningitis, and he turned to the Web to do research on the disease. What he discovered helped turn the fledgling analytics firm around. It was 2008, the economic crisis was beginning to take its toll, and Tynt's previous efforts to make a mark by analyzing social network data was beginning to fizzle.

"We were doing a project with social networks, but then the customers we were working with all got eaten by Facebook," recalls Tynt CEO Derek Ball, who says the experience helped Tynt get "very good at operating in the white space between the user and the publisher."

The epiphany came after MacKenzie spent a night surfing the Web for the meningitis information. He told his Tynt colleagues that he had gotten his results from hundreds of different Web sites. Ball recalls him saying, "Wouldn't it be great if when I cut and pasted something there was a way that could append where it came from?" Ball says that's when the Tynt team had its light-bulb moment: "What if we could do that and tell publishers what content people were copying and pasting, and how much that was drawing people to them? What if the content became a breadcrumb trail back to where the content came from? So we started looking at cut-and-paste and we found there was no research on it at all. No one had ever tried to study cut-and-paste rates."

Ball says the initial reaction was that there had to be one of two reasons why there was no research on it: "Either people had looked at it and the data just wasn't very relevant, or nobody had thought of doing it."

To find out how relevant it might be, the Tynt team developed an application to track the cut-and-paste rates of publishers who would install it on their pages. Initially, they deployed it on a Tynt engineer's Web site.

It was Christmas 2008, and that site - a guide to restaurant menus and contact information - was generating a tremendous amount of activity from users copying and pasting menu listings, addresses and phone numbers.

"We were astonished by the rates," Ball recalls, adding that the team was encouraged enough to take the beta version out for a real test drive: they installed it on Tynt investor Guy Kawasaki's blog. "That's when we found out that as much as six percent of the page loads on his site were coming from people cutting and pasting something," he says, "and that's when we realized that this data could be very useful to publishers who wanted to understand what was going on with people cutting and pasting their content."

Tynt began tagging URLs of publishers who wanted to understand cut-and-paste behavior, how it generated traffic back to their pages and where it was coming from.

The answer was surprising. While conventional online industry wisdom would indicate that most of the traffic comes from social platforms like Facebook and Twitter, Ball says the data disproves that: "The biggest social network in the world is actually email, and it is an uncontrolled, peer-to-peer network that nobody knows about."

Specifically, about 70 percent of all page loads resulting from copied and pasted links are generated by users pasting the links into emails and forwarding them to one or more other people.

Facebook is no slouch either, but it pales in comparison to conventional email communication. Ball estimates that about 25 percent of all traffic generated by copying and pasting comes from people pasting on Facebook pages.

Twitter, by comparison, is hardly even a factor.

"Currently, we're only seeing Twitter as being about two percent of all page loads," Ball says, though he thinks that is an underestimation. The actual rate might be twice that, he says, because many publishers do not have their address tracking bar turned on for Twitter.

In fact, Ball says the biggest issue - and where a big part of Tynt's time and energy is spent - involves analyzing the "signal-to-noise" ratios around copying and pasting. There is so much information related to the way people copy and save or share information on the Web that much of Tynt's resources go into "filtering" the noise to understand what data is truly meaningful.

"The trick is finding what's valuable in the signal, and which signals to disregard," he says. For example, he notes, it isn't simply macros or keystrokes that enable users to copy and paste information, but other actions as well, including printing.

"We really want to understand what people are interested in, and frankly, printing something off of the Web may be as strong an indication of someone's interest as copying and pasting something," he says. "For us, it's a similar signal of intent."

Fundamentally, Tynt's methodology isn't that complicated. It provides some simple code to publishers who register to be tracked, enabling them to create anonymous cookies that track where a unique user is coming from, and whether the traffic came as a result of pasting a link or a short code, or whether it was something shared with the user via an email or Facebook.

Tynt has even created a testing facility, dubbed Tynt Labs, to understand how other technologies motivate people to link, read and share Web content. For example, it is in the process of evaluating the impact of offline QR codes placed in analog media like magazines, newspapers, catalogues and billboards.

Regardless of how that behavior manifests itself, Ball says the goal is the same: to help publishers and marketers understand what their users are most interested in - and what triggers users to grab content and share it with other people.

In fact, Tynt's tagline is literally "who's interested." And, with more than 600,000 publishers now registered to track how users interact with their content, Ball says Tynt is also "the largest interest graph on the Internet."

It's not just "Joe's blog" that has signed up for Tynt; big publishers like Conde Nast and Hearst are using it to figure out what content their users find most compelling.

In the nearly three years since these behaviors began being tracked, Ball says Tynt hasn't seen any fall-off in the rates of activity. On average, he says, publishers reap between two percent and six percent of total traffic from copy-and-paste behavior, but on the high end copying and pasting can generate more than half of a site's total traffic. This is especially true for something like a travel site, as many users copy and paste listings of hotels, rates and other information before they book their trips.

So the big question is: with so many major publishers copying and pasting Tynt's tracking codes onto their pages, how come you have probably never heard of them? Don't worry, you're not alone. Last spring, during MediaPost's Social Media Insider Summit, while presenting data showing that Tynt is now the eighth largest Internet data collector - ranked just below AOL and ahead of Audience Science and Rubicon Project - Lou Kerner, an analyst with Wedbush Securities, asked how many of the Insiders had ever heard of it. No hands were raised.

To be fair, Ball says Tynt had been focusing primarily on publishers and only recently began pitching Madison Avenue directly. The data is just as compelling for brands as it is for content publishers, he says.

For example, Tynt's data reveals that, compared with other major automobile manufacturers' Web sites, Ford's are doing a much better job engaging consumers online.

"Ford is killing it right now, and Honda is too," Ball says, adding, "Wouldn't you like to know why?"

After reading this story, chances are that someone from one of Ford's or Honda's agencies - or those of their competitors - might find it interesting enough to copy, paste and forward it along to someone who does.